In the last year, we have focused on three specific events that occur during the life cycle of the model organism Bacillus subtilis in order to study how proteins localize within a cell and how they subsequently assemble into larger structures. First, we have examined the localization and assembly of the protein DivIVA during cell division. Second, we have studied how the peptide SpoVM orchestrates the timing of assembly of two different supramolecular structures during the developmental program of spore formation. Third, we have studied how the protein SpoIVA utilizes ATP to form stable filaments that envelope the maturing spore. During bacterial cell division, the tubulin homolog FtsZ assembles into a ring at mid-cell and constricts the plasma membrane during cytokinesis. In B. subtilis, this constriction is inhibited by a complex of three proteins whose subcellular localization depends on DivIVA. In the last year, using deconvolution fluorescence microscopy, we have demonstrated that DivIVA preferentially localizes to midcell, and secondarily to hemispherical poles of the rod-shaped bacterium. By reconstructing Z-stacks of deconvolved images, we have seen that DivIVA assembles into rings at mid-cell and patches at the poles. By examining time-lapse images of actively growing cells, we have determined the kinetics of DivIVA ring assembly and the stability of these rings once they have formed. However, DivIVA also forms patches elsewhere in the cell. How does one protein form two distinct structures? Again, by using time-lapse techniques, we observed that patches form when DivIVA rings collapse during cell separation;that is, when the architecture of the cell changes. Finally, in collaboration with Joseph Pogliano at UCSD, by employing a super resolution technique called structured illumination microscopy, we observed that DivIVA actually forms two adjacent rings on either side of nascent division septa. Thus, DivIVA represents an unusual example of a protein that assembles into a higher order structure, whose shape is determined by a pre-existing template: in this case, the curvature of the plasma membrane as defined by the shape of the cell. Taken together with our previous finding that DivIVA localizes by detecting negative membrane curvature, we have proposed that DivIVA localizes quickly to invaginating membranes (nascent cell division sites) and recruits FtsZ polymerization inhibitors in order to prevent aberrant cell division immediately adjacent to recent cell division sites. We are currently preparing a manuscript describing these results to submit for publication. When B. subtilis is faced with starvation conditions, it initiates a developmental program called sporulation, whereby it elaborates a dormant cell type called a spore, which stores the cells genetic material and protects it from environmental insults. During sporulation, the rod-shaped B. subtilis forms a spherical internal organelle called the forespore. The forespore is eventually encased in two concentric circles: a proteinaceous outer shell called the coat, and an inner shell made of peptidoglycan called the cortex. During sporulation, initiation of cortex assembly is dependent on the proper initiation of the coat, suggesting that a developmental checkpoint exists which links the morphogenesis of both structures. We have studied the role that a small peptide (called SpoVM) plays in orchestrating the assembly of the coat and cortex. Previously, we had shown that SpoVM detects the convex membrane curvature of the forespore to preferentially localize to that site and that SpoVM is directly or indirectly responsible for the subsequent recruitment of proteins that comprise the coat. In the last year, we have identified two amino acids in SpoVM whose substitution specifically disrupts cortex assembly, but not coat assembly, thereby separating the coat assembly function of SpoVM from its cortex assembly function. By employing genetic techniques we identified two other proteins that participate with SpoVM in initiation cortex assembly: a coat protein called SpoIVA that directly interacts with SpoVM and forms the basement layer of the coat, and a hitherto unidentified short protein encoded by a previously unannotated open reading frame. Current efforts are directed towards characterizing this new protein and in constructing a model whereby the tiny peptide SpoVM coordinates the morphogenesis of the spore coat and cortex. Proper localization of SpoVM to the surface of the forespore marks this surface as the site for the future localization of proteins that form the coat. The first coat protein that is recruited by SpoVM is SpoIVA which, as mentioned above, forms the basement layer of the coat. Our previous studies had shown that SpoIVA binds ATP via a Walker A motif and subsequently hydrolyzes the bound ATP, and we have shown that both binding and hydrolysis are required for the polymerization of the protein in vitro into long filaments. Our hypothesis is that ATP hydrolysis results in a conformational change in SpoIVA that promotes and stabilizes filament assembly, and that these filaments eventually encircle the forespore and provide a platform on top of which the rest of the coat forms. In order to test this model, we wondered if we could isolate variants of SpoIVA that are able to bind, but not hydrolyze ATP. Recently, we have identified Walker A and Walker B motifs in SpoIVA, which are typically implicated either directly or indirectly in ATP hydrolysis, but not ATP binding. Disruption of either motif resulted in drastic reductions of sporulation efficiency of cells producing these proteins, and purified SpoIVA harboring a disrupted Walker A motif displayed reduced ATP hydrolysis activity. We are currently establishing an assay for measuring polymerization of purified SpoIVA in order to correlate ATP hydrolysis activity with polymerization.